Isn't a rather good deal of color from feathers a result of "structural color", rather than pigmentation? I'd be curious if fossilized feathers could ever, in theory, preserve enough microscopic detail to guess at that.